Task-Oriented Repetitive Training After Stroke: Neuroplasticity, Dose and the Role of Robotics

After a stroke, the brain retains a remarkable capacity for reorganisation. This capacity, known as neuroplasticity, constitutes the scientific foundation on which modern neurological rehabilitation is based. However, for that reorganisation to translate into functional improvements, evidence suggests that a sufficient stimulus is needed in the form of repetitive, intensive and task-oriented practice. This article analyses the principles of neuroplasticity, the importance of dose and intensity in stroke rehabilitation, and how robotic technology can contribute to optimising these parameters.

1. Principles of Neuroplasticity After Stroke

Neuroplasticity refers to the capacity of the central nervous system to modify its structure and function in response to experience and training. After brain injury such as stroke, this property acquires fundamental clinical relevance, as it enables reorganisation of damaged neural networks or the assumption of functions by unaffected brain areas.

Hebbian Learning: The Basis of Reorganisation

One of the central principles of neuroplasticity is Hebbian learning, formulated by Donald Hebb in 1949 and habitually summarised as “neurons that fire together, wire together.” This principle establishes that the synchronous and repeated activation of synaptic connections strengthens those connections, consolidating specific neural circuits.¹

In the context of stroke rehabilitation, Hebbian learning implies that the repeated practice of a movement or functional task can contribute to strengthening the neural pathways associated with that function. The more frequent and consistent the activation of a motor circuit, the greater the probability that circuit will stabilise and consolidate.

Use-Dependent Plasticity

Closely related to Hebbian learning, the concept of use-dependent plasticity indicates that cortical reorganisation is conditioned by active motor experience. Classic studies by Nudo et al. (1996) in primate models demonstrated that, following cortical motor injury, intensive practice of tasks with the affected limb induced an expansion of the cortical representations of the trained muscles.²

Conversely, lack of use —as occurs when the patient compensates with the unaffected limb— can lead to further reduction of cortical representations, a phenomenon known as learned non-use, extensively described by Taub et al.³ Evidence therefore suggests that effective rehabilitation requires intensive and specific activation of the functions to be recovered.

Time Window and Enhanced Sensitivity

There is a period of heightened sensitivity to neuroplasticity after stroke, especially during the first weeks and months. During this phase, sometimes called the “critical period” or “enhanced neuroplasticity window,” the brain shows a greater response to rehabilitation stimuli.⁴ However, brain plasticity is not limited to this phase: evidence indicates that the capacity for reorganisation persists in chronic phases, although potentially with a smaller response more dependent on stimulus intensity.⁵

2. Dose and Intensity: How Many Repetitions Are Needed?

If neuroplasticity depends on repeated practice, a key clinical question is: how much practice is enough? The answer, although not yet fully defined, consistently points towards repetition volumes far higher than those typically achieved in conventional clinical practice.

Evidence on Dose in Animal Models

Studies in animal stroke models have provided indicative data on the practice dose required. MacLellan et al. (2011) observed that rats with cortical lesions performing at least 400 daily reaching repetitions showed significantly greater functional recovery than those with lower volumes.⁶ While not directly extrapolable to humans, these data establish a reference framework underscoring the importance of practice quantity.

The Gap Between Evidence and Clinical Practice

One of the most relevant findings in stroke rehabilitation research is the enormous discrepancy between the practice doses recommended by evidence and those actually achieved in conventional sessions. Lang et al. (2009) documented that, during typical occupational therapy and physiotherapy sessions in stroke rehabilitation, patients performed a median of only 32 functional upper limb repetitions per session.⁷

This figure contrasts starkly with the hundreds or thousands of repetitions that neuroscience evidence suggests are necessary to induce significant plastic changes. Birkenmeier et al. (2010) noted that increasing repetitions to 300 or more per session within a task-oriented programme produced additional functional improvements.⁸

“Task-oriented repetitive practice can contribute to driving brain reorganisation, but dose matters. Typical repetition volumes in conventional clinical practice may fall well below the threshold needed to induce significant neuroplastic changes.”

Intensity and Task Specificity

Dose does not refer solely to the gross number of repetitions, but also to task specificity. The principle of training specificity establishes that the benefits of practice are greater when the trained task closely resembles the functional activity to be improved.⁹ For example, repetitive practice of reaching and grasping objects may be more effective for upper limb function than isolated analytical exercises of a single muscle group.

French et al. (2016), in a Cochrane review, concluded that task-oriented repetitive training can contribute to improving upper limb function and gait ability in people after stroke, although evidence quality was moderate and more data on optimal dose were needed.¹⁰

3. Evidence on Repetition and Functional Outcomes

The relationship between practice volume and functional outcomes has been investigated in multiple clinical trials and systematic reviews. Available data suggest that greater practice intensity tends to be associated with better outcomes, although the relationship is not linear and important moderating factors exist.

Upper Limb Repetitive Training

The Cochrane review by Pollock et al. (2014) on upper limb function interventions after stroke identified task-oriented repetitive training as one of the interventions with favourable evidence, especially when applied with sufficient doses.¹¹ Lohse et al. (2014), in a dose-response meta-analysis, found that greater therapy time was associated with better functional outcomes, with a relationship showing no clear ceiling effect within the studied ranges.¹²

Gait Training

Regarding gait rehabilitation, evidence also supports the importance of practice volume. Hornby et al. (2020), in the American Physical Therapy Association clinical practice guidelines, recommended that stroke patients who can walk should accumulate as much over-ground walking practice as possible, at moderate to high cardiovascular intensity.¹³

Moore et al. (2010) demonstrated that treadmill gait training sessions with a greater number of steps were associated with greater improvements in gait speed and ambulation capacity.¹⁴ The key lies not only in walking more, but in doing so with as normalised a gait pattern as possible, linking intensity with task specificity.

The Role of Feedback

Repetition alone may not be sufficient. Feedback —both intrinsic (sensory) and extrinsic (provided by the therapist or technology)— is a fundamental component of motor learning. Subramanian et al. (2010) found that performance feedback, combined with repetitive practice, enhanced gains in upper limb function.¹⁵ Robotic devices, as we shall see, can offer objective and real-time feedback, which can contribute to improving the quality of each repetition.

4. The Role of Robotics in Increasing Repetitions

If evidence suggests hundreds of repetitions per session are needed and conventional clinical practice rarely achieves these figures, a natural question arises: how can we safely and sustainably increase practice volume? Robotic technology presents itself as a valuable complement to address this challenge.

Advantages of Robot-Assisted Rehabilitation

Robotic devices for neurorehabilitation, such as the robotic gait training system and the upper limb rehabilitation robot, offer several features that can contribute to optimising practice dose:

• Greater volume of repetitions: by providing mechanical assistance to movement, robotic devices allow the patient to perform a significantly greater number of repetitions per session. For example, a robotic gait training system can facilitate hundreds or thousands of gait cycles in a single session, a volume difficult to achieve in conventional manual therapy.¹⁶
• Adaptive assistance: modern robotic systems incorporate control algorithms that adjust the level of assistance to the patient's effort, providing only the help needed. This “assist-as-needed” principle promotes active participation, a key factor for neuroplasticity.¹⁷
• Objective feedback: robotic devices record detailed biomechanical data —force, range of motion, symmetry, speed— allowing the clinical team to objectively monitor progress and adjust the treatment programme accordingly.
• Safety and postural support: especially in patients with severe motor deficits, robotics allows intensive practice with biomechanical support that reduces the risk of falls or inappropriate compensations.

Evidence: Robotics vs. Conventional Therapy

Mehrholz et al. (2020), in their Cochrane review on electromechanical-assisted gait training after stroke, concluded that the combination of robotic gait training with conventional physiotherapy can increase the probability of achieving independent gait compared with physiotherapy alone, especially in patients who could not walk at the start of treatment.¹⁶

Regarding the upper limb, Mehrholz et al. (2018) found that robot-assisted rehabilitation can contribute to improving arm function and activities of daily living, although effects were modest and depended on the protocol used.¹⁸ The authors emphasised that robotics should be used as a complement to a conventional rehabilitation programme, not as a substitute.

“Robotic technology does not replace the physiotherapist's intervention, but can contribute to amplifying its impact by allowing practice volumes that would otherwise be difficult to achieve.”

5. Practical Application at GNeuro

At GNeuro, a robotic neurorehabilitation centre located in Ourense, these principles of neuroplasticity and task-oriented repetitive training are integrated into every rehabilitation programme. Our approach combines the clinical expertise of professionals specialised in neurological damage with state-of-the-art robotic technology.

Technology at the Service of Intensity

The centre has a robotic gait training system for locomotion training and an upper limb rehabilitation robot for upper limb rehabilitation. These devices allow:

• Sessions with a high volume of repetitions, oriented towards functional tasks such as gait or reaching and grasping.
• Individual adjustment of assistance, promoting active patient participation according to the “assist-as-needed” principle.
• Recording objective data from each session to monitor progress and adapt the therapeutic programme.

Individualised Programmes

Each patient at GNeuro receives a comprehensive initial assessment by a physiotherapist specialised in neurorehabilitation, in which functional goals are defined, motor status is assessed and the suitability of robotic rehabilitation as a complement to the individualised programme is determined.

Robotic therapy is integrated within a programme that also includes conventional physiotherapy, speech therapy when needed and continuous follow-up. Evidence suggests that it is this combination of approaches, not an isolated device, that can offer the best outcomes.

Commitment to Evidence

At GNeuro, clinical decisions are based on the best available scientific evidence. Selection of protocols, doses and interventions is continuously reviewed in light of the literature and updated clinical practice guidelines.

Frequently Asked Questions